It is common knowledge that party support in England is strongly polarized along class lines. But how strongly? And is the degree of class polarization increasing, decreasing, or approximately constant? It is to these questions that this article is devoted.

Rapidly changing economic conditions and the acceptance by governments of the responsibility for those conditions have together provided one of the most volatile and perplexing policy contexts for governments in the post-war era. The volatility of economic policy is inherent in its potential for change over the short run. It can take several years before changes are approved and implemented in order to deal with failures in social welfare programmes, transport services, or redistributive taxation provisions. In contrast, the instruments of economic policy may be altered on a daily, monthly, quarterly or yearly basis as conditions change.

As a consequence of Truman's Point Four Program of 1949, the provision of economic assistance to independent low-income countries became the official policy of the United States. In the early and mid-1950s economic development assistance, though growing, received relatively little attention as the Korean war turned the USA's foreign aid in the direction of military assistance. However, by the start of the 1960s, the transfer of economic assistance from high-income to low-income countries had developed into an institutionalized relationship. Economic assistance was clearly distinguished from military assistance and administered separately; the USA's monopoly of aid was decreasing as other high-income countries, partially under the USA's pressure, were establishing their own aid programmes; and the volume of aid and the number of recipients were increasing.

In this article we take advantage of a two-wave panel study of two biologically-linked generations in order to examine some prevalent conceptions about the persistence of political orientations. The nature of the study design, outlined below, is particularly suited to addressing questions about individual-level continuities as they are affected by life-cycle, generational and historical processes. Our present discussion is geared to the individual level rather than to the aggregate level of continuity and change. Although many of the terms used are the same, and although it is difficult to discuss the one level without recourse to the other, the purposes and the approaches are fundamentally different. Aggregate analysis concerns itself with net movements and with the directionality of these movements. Individual analysis, as used here, concerns itself with the magnitudes of individual-level movements and has only a secondary interest in the direction of these movements.

For political scientists and pollsters the way the individual voted on previous occasions provides an important source of data. In the absence of longitudinal studies, recall of past vote tends to be taken as equivalent to actual vote cast. How accurate is such recall? How far does accuracy decrease with time, where recall concerns not one, but two, previous elections? How far do errors introduce a systematic bias in the conclusions drawn from such data?

In a recent article (‘On Joining Interest Groups: An Empirical Consideration of the Works of Mancur Olson Jr.’, this Journal, VI (1976), 257–72), David Marsh has attempted to test Mancur Olson's hypotheses concerning the rationale for individual or corporate membership of interest groups by using data from a study of the Confederation of British Industry. It is the purpose of this Note to defend Professor Olson's broad theoretical approach from some of Dr Marsh's criticisms. Olson applies the insights of established economic analysis to political theory; it is not his methodology which represents an innovation, but rather the use to which it is put. A defence of Olson's approach is, then, of necessity to some extent a more general defence of economic analysis.

Shackleton's comment on my earlier article seems to me to raise a number of interesting questions. He is very critical of my methodology and suggests alternative ways of testing Olson's model. He places most emphasis on the need to consider aggregate data on the changing patterns of interest group membership over time. In this short note I do not intend to reply at length to Shackleton's criticisms. Instead, I shall re-examine Olson's analysis using aggregate data on the changing patterns of CBI membership in the last decade.